Re: Pressing ? does not allow window scrolling through the list of info commands

[Eli Zaretskii <eliz@gnu.org>]

> Date: Fri, 07 Jul 2023 22:40:06 +0300
> From: Eli Zaretskii <eliz@gnu.org>
> 
> > From: Yuri Khan <yuri.v.khan@gmail.com>
> > Date: Sat, 8 Jul 2023 02:34:56 +0700
> > Cc: help-gnu-emacs@gnu.org
> > 
> > On Sat, 8 Jul 2023 at 02:06, Eli Zaretskii <eliz@gnu.org> wrote:
> > 
> > > > ‘help-mode-map’ has ‘q’ bound to ‘quit-window’ and that’s the minimal
> > > > fuss way to return from the help window. It also would help build
> > > > habits that work with a wide set of buffers in Emacs.
> > >
> > > And how will we tell the user that 'q' quits?  In the echo-area,
> > > perhaps?
> > 
> > Echo area, or header line, or mode line, whatever.
> 
> Which is what we do now about SPC, and that is deemed not good enough.
> How is showing 'q' there suddenly better?
> 
> > > > As it is, the user has to build three sets of habits […]
> > >
> > > You exaggerate the problem.  A typical user of computers these days
> > > needs much more than 3 sets of habits for similar actions.  And
> > > scrolling with SPC in Emacs is quite a widely-used paradigm.
> > 
> > Users who know about scrolling with Space also expect Shift+Space to
> > scroll in reverse. Users also *really* expect to be able to scroll by
> > single lines. Help that hides as soon as you try to scroll it is not
> > very helpful.
> 
> I disagree.

As an aside, this is a typical Emacs incarnation of "the road to hell
is paved with good intentions": we take an obscure minor feature that
5 min ago no one even knew it existed and which bothered no one, and
suddenly it is a huge usability issue, worthy of inventing novel ways
of user interactions, new faces, new header-line displays (which, btw,
steal one screen line used for precious Help display), and whatnot.
And then we are surprised that Emacs doesn't become more stable as the
years go by.

Will we ever learn?